Near x=0 the equation x2y′′+xy′−94y=0, with independent variable x and y′=dy/dx, has the solutions x2/3 and x−2/3, and the similar equation x2y′′+xy′=0 has the solution lnx. No power series in whole powers of x behaves like these near 0, so the ordinary-point method cannot find them. Which equations still allow a series solution at a singular point, and what form must it take? This lesson classifies regular singular points, finds the leading power from the indicial equation, and shows when the second solution needs a logarithm. Throughout, the work is on the interval x>0, where xr=erlnx and lnx are defined without any choice of branch.
Key terms
Regular singular point
Definition Let x0 be a singular point of the normalized equation
y′′+p(x)y′+q(x)y=0,so that p or q is not analytic at x0. The point x0 is a regular singular point if both (x−x0)p(x) and (x−x0)2q(x) are analytic at x0, once their values at x0 are defined by their limits. Otherwise x0 is an irregular singular point. For a regular singular point at x0=0, multiplying the normalized equation by x2 gives the standard form
x2y′′+xP(x)y′+Q(x)y=0,where P(x)=xp(x) and Q(x)=x2q(x) are analytic at 0. A regular singular point at another x0 moves to 0 under the substitution s=x−x0.
In plain words At a regular singular point, p has at worst a simple pole and q at worst a double pole. That is the same strength of singularity as in a Cauchy-Euler equation, and it is mild enough for a modified series.
Example For x2y′′+xy′−94y=0 the normalized coefficients p(x)=1/x and q(x)=−4/(9x2) are not analytic at 0, but xp(x)=1 and x2q(x)=−94 are constants. So 0 is a regular singular point with P=1 and Q=−94. Bessel’s equation x2y′′+xy′+(x2−ν2)y=0 is regular singular at 0 as well, with P=1 and Q(x)=x2−ν2. By contrast, x3y′′+y=0 has q(x)=1/x3, and x2q(x)=1/x is not analytic at 0, so 0 is an irregular singular point.
Frobenius series
Definition At a regular singular point 0, a Frobenius series is a trial solution on x>0 of the form
y=xrn=0∑∞anxn=n=0∑∞anxn+r,with a0=0, where the exponent r is to be found and may be fractional or negative. The condition a0=0 makes xr the actual leading power. The theorem of Frobenius says when such a solution exists. Suppose the series of P and Q about 0 converge for ∣x∣<ρ, and the roots r1≥r2 of the indicial equation below are real. Then there is a solution y1=xr1∑anxn with a0=1 whose series converges at least for ∣x∣<ρ, so y1 solves the equation on 0<x<ρ. If r1−r2 is not an integer, there is a second solution of the same form with exponent r2. On x<0 the same statements hold with xr replaced by ∣x∣r.
In plain words Let the equation choose a leading power xr first, then correct it with an ordinary power series. Close to 0 such a solution behaves like a constant times xr.
Example For x2y′′+xy′−94y=0 the coefficients P and Q are constant, and the Frobenius series reduce to single powers. For any power,
x2(xr)′′+x(xr)′=r2xr,so y=xr is a solution exactly when r2=94, which gives x2/3 and x−2/3. The singular point does not accept arbitrary initial values, so data are imposed at an ordinary point such as x=1. For y=Ax2/3+Bx−2/3 the conditions y(1)=3 and y′(1)=32 read A+B=3 and 32(A−B)=32, so A=2, B=1 and
y=2x2/3+x−2/3.Indicial equation
Definition Substitute a Frobenius series into x2y′′+xP(x)y′+Q(x)y=0, where P(x)=∑Pkxk and Q(x)=∑Qkxk. The lowest power xr has the coefficient a0F(r), with
F(r)=r(r−1)+P(0)r+Q(0).Since a0=0, the exponent must satisfy the indicial equation F(r)=0, and its roots are called the indicial roots, or exponents, at the singular point. For n≥1 the coefficient of xn+r gives
F(r+n)an=−k=0∑n−1cn,kak,cn,k=(k+r)Pn−k+Qn−k.This determines an from the earlier coefficients whenever F(r+n)=0. At an index where F(r+n)=0 the equation for an cannot be solved in the usual way, and that is where the difficulties with the second solution come from.
In plain words Very close to 0 only the leading terms matter, and there the equation acts like the Cauchy-Euler equation x2y′′+P(0)xy′+Q(0)y=0. Its exponents are the possible leading powers of the true solutions.
Example Multiplying xy′′+y=0 by x gives x2y′′+xy=0, so P=0 and Q(x)=x, and 0 is a regular singular point. The indicial equation is r(r−1)=0, with roots 1 and 0. For r=1 the only nonzero term on the right is the one with Q1=1, and F(n+1)=n(n+1), so the recurrence is an=−an−1/(n(n+1)). With a0=1 it gives
y1=n=0∑∞n!(n+1)!(−1)nxn+1=x−2x2+12x3−⋯,which converges for every x, since P and Q are polynomials.
Roots differing by an integer
Definition Let the indicial roots be real with r1−r2=N for a positive integer N. The larger root always gives a Frobenius solution y1, because F(r1+n)=0 for every n≥1. For the smaller root the recurrence reaches n=N with F(r2+N)=F(r1)=0, so the equation for aN reads 0⋅aN=SN, where SN is the right side computed from a0,…,aN−1. If SN=0, no Frobenius solution with exponent r2 exists, and the second solution contains a logarithm. If SN=0, every value of aN works, and r2 gives a second Frobenius solution with no logarithm. Changing aN only adds a multiple of y1. When r1−r2 is not an integer, both roots give Frobenius solutions, and they are independent because their leading powers differ. An integer difference is therefore a reason to check the equation at n=N, not a proof that a logarithm appears.
In plain words When the two leading powers are a whole number apart, the series that starts at the lower power eventually reaches the higher one. At that step the equation either lets it pass or forces a logarithm.
Example For x2y′′+xy′−41y=0 the roots are 21 and −21, so N=1, yet x1/2 and x−1/2 both solve it. Here P and Q are constant, so S1=0. For xy′′+y=0 the roots 1 and 0 also differ by 1, but with r2=0 the equation at n=1 is F(1)a1=−Q1a0, that is 0⋅a1=−a0. This is impossible because a0=0, so that equation has no second Frobenius solution, and its second solution must contain lnx.
Repeated indicial root
Definition The indicial root r1 is repeated when F(r)=(r−r1)2. Then F(r1+n)=n2=0 for every n≥1, so the recurrence determines every coefficient from a0, and every Frobenius solution with exponent r1 is a constant multiple of y1. A second independent solution therefore cannot be a Frobenius series. It has the form
y2=y1lnx+xr1n=1∑∞bnxnon 0<x<ρ, so in this case the logarithm is always present.
In plain words A double root supplies only one leading power, so the second solution has to come from somewhere else. It comes from a factor lnx, just as a double root of a Cauchy-Euler equation gives xr1 and xr1lnx.
Example The equation x2y′′+xy′=0 has P=1 and Q=0, so F(r)=r2 and 0 is a double root. The constant y1=1 is a solution. With v=y′ the equation becomes xv′+v=0, so v=C/x and a second solution is y2=lnx. Its derivatives 1/x and −1/x2 give x2⋅(−1/x2)+x⋅(1/x)=0. The Wronskian W(1,lnx)=1⋅(1/x)−0⋅lnx=1/x is nonzero on x>0, so 1 and lnx form a fundamental set there. In the form above, every bn is zero.
Logarithmic second solution
Definition Let 0 be a regular singular point with real indicial roots r1≥r2, and let the series of P and Q converge for ∣x∣<ρ. If the roots are equal, a second solution on 0<x<ρ has the form shown for a repeated indicial root, with the coefficient of y1lnx equal to 1. If r1−r2 is a positive integer, a second solution on 0<x<ρ is
y2=Cy1lnx+xr2n=0∑∞bnxnwith b0=0, where the constant C is zero exactly when the smaller root gives a Frobenius solution. In both cases the series converge at least for ∣x∣<ρ, and y1 and y2 form a fundamental set on 0<x<ρ. The coefficients are found by substituting the form into the equation. Alternatively, on an interval where y1 has no zeros, reduction of order gives
y2=y1∫y12e−∫pdxdx,and the logarithm appears when the integrand contains a term 1/x. On x<0 the same forms hold with ∣x∣ in place of x.
In plain words When the second leading power is missing or blocked, the equation makes up the shortfall with lnx times the first solution, plus a correcting series.
Example For x2y′′−xy′+y=0 the indicial polynomial is r(r−1)−r+1=(r−1)2, so r=1 is repeated and y1=x. Here p(x)=−1/x and e−∫pdx=x, so reduction of order gives
y2=x∫x2xdx=xlnx,with the logarithm coming from ∫dx/x. For xy′′+y=0, where the smaller root 0 is blocked, substituting the second form with b0=1 and the free choice b1=0 gives
y2=−y1lnx+1−43x2+367x3−⋯,with y1 from the indicial equation card. The constant is forced, since the constant term of xy2′′+y2 works out to C+b0, so C=−1.
Check yourself
- Classify the point x=0 for x2y′′+3xy′+(1+x)y=0 and for x3y′′+xy′+y=0. For the regular singular one, find the indicial roots and say what they predict about the second solution.
Show the answer
For the first equation p(x)=3/x and q(x)=(1+x)/x2, so 0 is singular, but xp(x)=3 and x2q(x)=1+x are analytic, so 0 is a regular singular point with P=3 and Q(x)=1+x. The indicial polynomial is
r(r−1)+3r+1=(r+1)2,
so −1 is a repeated root. Up to a constant factor there is one Frobenius solution with exponent −1, which works out to y1=x−1∑(−1)nxn/(n!)2, and the second solution must contain y1lnx. For the second equation p(x)=1/x2, and xp(x)=1/x is not analytic at 0, so 0 is an irregular singular point and the Frobenius theory does not apply there.
- Find the indicial roots of 2x2y′′+3xy′−(1+x)y=0 at 0. How many Frobenius solutions are guaranteed? Find the first three coefficients of each with a0=1.
Show the answer
Dividing by 2 gives the standard form with P=23 and Q(x)=−21(1+x), so
F(r)=r(r−1)+23r−21=(r+1)(r−21).
The roots 21 and −1 differ by 23, which is not an integer, so there are two independent Frobenius solutions, and since P and Q are polynomials their series converge for all x. In the original equation the coefficient of xn+r is (2(n+r)−1)(n+r+1)an−an−1. For r=21 this gives an=an−1/(n(2n+3)), so a1=51, a2=701 and
y1=x1/2(1+5x+70x2+⋯).
For r=−1 it gives an=an−1/(n(2n−3)), so a1=−1, a2=−21 and
y2=x−1(1−x−2x2−⋯).
- The indicial roots of xy′′+2y′+xy=0 at 0 differ by an integer, and a classmate concludes that the second solution must contain lnx. Check this against the recurrence and identify both solutions.
Show the answer
Multiplying by x gives x2y′′+2xy′+x2y=0, so P=2 and Q(x)=x2. The indicial equation r(r−1)+2r=r(r+1)=0 has the roots 0 and −1, with N=1. The coefficient of xn+r is F(n+r)an+an−2, with a−2=a−1=0. For r=−1 and n=1 this reads F(0)a1=0, that is 0⋅a1=0, so there is no obstruction and a1 is free. The classmate is wrong. With a1=0 the recurrence an=−an−2/(n(n−1)) makes every odd coefficient zero and gives a2=−21, a4=241 and in general a2k=(−1)k/(2k)!, so
y=k=0∑∞(2k)!(−1)kx2k−1=xcosx.
The root 0, with an=−an−2/(n(n+1)) and a1=0, gives a2k=(−1)k/(2k+1)! and the solution sinx/x. A direct check confirms both, because the equation can be written (xy)′′+xy=0, so xy is a combination of cosx and sinx. Neither solution has a logarithm.